Low-Correlation Superimposed Pilot Grant-Free Massive Access for Satellite Internet of Things

Satellite Internet of Things (S-IoT) with low Earth orbit satellites has become an effective solution for providing global coverage for massive machine type communications (mMTC). Considering that the massive user equipments covered by the S-IoT are periodically activated and dominated by short packet communications, the pilot collision has become a challenging problem due to the limited length and number of pilot sequences. In this paper, we propose a low-correlation superimposed pilot grant-free massive access (LSP-GFMA) scheme, where a low-correlation-zone periodic sequence (LPS) is designed for the superimposed pilot (SP) structure. Our LPS can maintain low cross-correlation with random non-orthogonal shifts compared with the conventional Zadoff-Chu sequence (ZCS), which can alleviate pilot collision while ensuring high spectral efficiency. In addition, we propose an iterative channel estimation based on Kaczmarz algorithm to attain accurate channel state information for the SP structure with low complexity. Then, we derive the theoretical expressions of access failure probability (AFP) and achievable throughput for our LSP-GFMA scheme under the shadowed-Rician fading channel. Simulation results validate the accuracy of our theoretical derivations, and demonstrate that our LSP-GFMA scheme with LPS can achieve lower AFP and higher achievable throughput than that with ZCS, and also outperforms the state-of-the-art schemes.